Methods for measuring position of phantom sound sources?

[Thomas_A]

In another forum there is a discussion whether and how stereo systems differ in terms of how phantom images are localised when you move seating position in relation to speakers. In some systems the image (e.g. center dialogue) tilts to the nearest speaker when moving towards that, in other systems the center remains in center even if you sit in front of one of the speakers.

This question is whether this phenomenon is measurable?

(Perhaps there is a thread about this subject alredy, but I have not found one.)

Edit: I found one thread..., this one can be deleted.

 

[dasdoing]

Edit: I found one thread..., this one can be deleted.

which?

 

[Dumdum]

In another forum there is a discussion whether and how stereo systems differ in terms of how phantom images are localised when you move seating position in relation to speakers. In some systems the image (e.g. center dialogue) tilts to the nearest speaker when moving towards that, in other systems the center remains in center even if you sit in front of one of the speakers.

This question is whether this phenomenon is measurable?

(Perhaps there is a thread about this subject alredy, but I have not found one.)

Edit: I found one thread..., this one can be deleted.

What your describing isn’t really possible and won’t ever give a great centre image without a centre speaker, and even then you will have quite bad comb filtering at one side or another if getting near to be sat in front of a speaker

Timing will be off massively and the centre will move to one side or the other despite claims by others to the contrary in a two speaker setup, also levels will further compound the issue

 

[Thomas_A]

which?

I can’t remember now but it was discussed in terms of binaural measurments.

 

[Thomas_A]

What your describing isn’t really possible and won’t ever give a great centre image without a centre speaker, and even then you will have quite bad comb filtering at one side or another if getting near to be sat in front of a speaker

Timing will be off massively and the centre will move to one side or the other despite claims by others to the contrary in a two speaker setup, also levels will further compound the issue

Well I am not sure I agree that measurements can’t be done but they are complicated. As Toole also mentioned in another thread there are more mechanisms relating to locking on positions that is more a brain and learning process. These are not possible to measure other than with human experiments.

 

[dasdoing]

it is possible in theory with a very controled off axis response and a steep tow-in. the center-LP sits off-axis to both speakers = reduced output. left-LP is farer from the right speaker but on-axis; now you could balance stuff out so he gets same output as center-LP (louder onaxis but bigger distance). the left speaker is nearer but now way offaxis; again now you could balance stuff out so he gets same output as center-LP.
all this would probably be possible with some form of beamforming tecnology

 

[LTig]

You just best me. Yep, such a toe in of lets say 45 degree can offer a wider sweet spot.

 

[Thomas_A]

If I get time I may do and upload some ”binaural” audio files of left-both-right speakers.

 

[dasdoing]

You just best me. Yep, such a toe in of lets say 45 degree can offer a wider sweet spot.

you could design a speaker for this though. it needs a flat off-axis with a certain (calculated) atenuation on purpouse.
if you take a conventional speaker for this you can only match part of the spectrum.

 

[Dumdum]

Well I am not sure I agree that measurements can’t be done but they are complicated. As Toole also mentioned in another thread there are more mechanisms relating to locking on positions that is more a brain and learning process. These are not possible to measure other than with human experiments.

I was referring to a set of speakers where you could sit off centre and they play a perfect centre also, a lack of time alignment will make massive comb filters regardless of response in higher freqs how high depends how big the listening triangle is and distance to speakers, amplitude is just half of the equation, I live in the car audio realm where we do sit in front of the drivers side speaker and there isn’t any nice way to get a good centre image with identical timing of left and right without an upmixer and a centre channel, it just won’t happen, it’s a romantic notion, but you can’t correct for timing and off axis response for more than one sweet spot ever for me

 

[Thomas_A]

I was referring to a set of speakers where you could sit off centre and they play a perfect centre also, a lack of time alignment will make massive comb filters regardless of response in higher freqs how high depends how big the listening triangle is and distance to speakers, amplitude is just half of the equation, I live in the car audio realm where we do sit in front of the drivers side speaker and there isn’t any nice way to get a good centre image with identical timing of left and right without an upmixer and a centre channel, it just won’t happen, it’s a romantic notion, but you can’t correct for timing and off axis response for more than one sweet spot ever for me

There is never a perfect phantom center off-axis but still enough differences between setups that should be measurable (excluding the psychoacoustics involved). The question is related to how such masurements should be done. Nothing more than that.

 

[Dumdum]

There is never a perfect phantom center off-axis but still enough differences between setups that should be measurable (excluding the psychoacoustics involved). The question is related to how such masurements should be done. Nothing more than that.

In that case you could measure phase and magnitude along with impulse response at various places and compare, that will give you timing and how far it is out and then repeat for hundreds of other speakers until you find a trend